Re: [CTPP] census dorm errors

Sounds a lot like what I observed with the 1990 Census block data for Winston-Salem/Forsyth County, North Carolina, primarily in recently developed areas. In that instance, I was comparing the census data to recently collected field data and current aerial photography. Where discrepancies existed, it was fairly easy to imagine that our data collectors could have missed units, however, there were many instances where the census indicated dwelling units in clearly undeveloped areas (no buildings). Well established older neighborhoods, appeared to be very accurate. The County-wide figures, likewise, were in fairly close agreement with our local data. This seems to be common based on my limited experience. Unfortunately, I have no direct experience with the 2000 census, and although I understand there are discrepancies with our 2000 data, we will probably not mount a serious challenge because of our negative experience with the 1990 challenge. Mark Wilkes, PE, AICP Director of Transportation Planning Chatham County-Savannah Metropolitan Planning Commission P.O. Box 8246, 110 East State Street Savannah, GA 31412-8246 wilkesm(a)thempc.org tel. (912) 651-1451 fax (912) 651-1480 tmank@tompkins-co .org To: ctpp-news(a)chrispy.net Sent by: cc: owner-ctpp-news@c Subject: [CTPP] census dorm errors hrispy.net 06/05/2001 10:42 AM I wanted to pass this on to the listserv: I work for the Tompkins County Planning Department in Upsate, NY. When I first looked at the Block level populations this year, I noticed numerous discrepancies, which turned out to be dorm population counts which were incorrectly georeferenced. We have 2 universities here in Ithaca, NY, Cornell University and Ithaca College. In our case, the entire dorm population for Cornell (5,700 students) was put in ONE BLOCK - not in the 14 blcoks where the dorms are actually located. For example, a block that had a 1990 population of 1100 now has a 2000 pop of 2 and a block that had a 1990 pop of 110 now has a 2000 pop of 5882. For Ithaca College, the dorm population was placed all in 1 block, not in the 4 blocks where the dorms are actually located. To make things more complicated, the Cornell Block discrepancies go across municipal boundaries (from the City of Ithaca to the Town of Ithaca) and across four of our current Legislative Districts [see attachement with PDF map]. When I brought this up to the Regional Census Geography dept in Boston, they were not that surprised and said they were seeing this all over the country. The New York State Data Center in Albany says they have had this complaint all over New York State. In any case, we have noted the dorm capacities and redistributed the block populations as we think they should be and will submit for a change through the Census Bureau's Question Resolution Program, which starts in July. We also have gone ahead with our 2001 Reapportionment for Tompkins County pending the Census Bureau's approval of the changes. My questions are: 1) How many of the MPOs are experiencing the same problem? 2) I am concerned that we can estimate the Census errors and change the Block populations, but ONLY THE CENSUS Bureau knows the exact number of students who were in those dorms and their ATTRIBUTES. Only the Census can move those attributes - i.e. those 5,700 student incomes and journey-to-work responses, etc. It seems to me that all the long form data that comes out next year will be skewed - unless they deal with this issue. Which will probably only happen if the issue is as wide spread as I have been made to believe. So what are you all seeing in this regard? Tom Mank Planning Analyst Tompkins County Planning Department The following section of this message contains a file attachment prepared for transmission using the Internet MIME message format. If you are using Pegasus Mail, or any another MIME-compliant system, you should be able to save it or view it from within your mailer. If you cannot, please ask your system administrator for assistance. ---- File information ----------- File: 00bloc~1.pdf Date: 19 Apr 2001, 5:42 Size: 398342 bytes. Type: Unknown (See attached file: 00bloc~1.pdf)